R Notebook

library(tidyverse)

Registered S3 methods overwritten by 'dbplyr':
  method         from
  print.tbl_lazy     
  print.tbl_sql      
── Attaching packages ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────── tidyverse 1.3.0 ──
✓ ggplot2 3.3.2     ✓ purrr   0.3.4
✓ tibble  3.0.4     ✓ dplyr   1.0.2
✓ tidyr   1.1.2     ✓ stringr 1.4.0
✓ readr   1.4.0     ✓ forcats 0.5.0
── Conflicts ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────── tidyverse_conflicts() ──
x dplyr::filter() masks stats::filter()
x dplyr::lag()    masks stats::lag()

df <- read.csv("../ESI 4011 ~ Smart Cities/yellow_tripdata_2020-01.csv")
head(df)

summary(df)

    VendorID              tpep_pickup_datetime         tpep_dropoff_datetime passenger_count trip_distance         RatecodeID    store_and_fwd_flag  PULocationID    DOLocationID    payment_type    fare_amount      
 Min.   :1.00    2020-01-09 16:53:00:     21   2020-01-12 00:00:00:     69   Min.   :0.00    Min.   :   -30.62   Min.   : 1.00    :  65441          Min.   :  1.0   Min.   :  1.0   Min.   :1.00    Min.   :-1238.00  
 1st Qu.:1.00    2020-01-06 08:13:00:     20   2020-01-11 00:00:00:     67   1st Qu.:1.00    1st Qu.:     0.96   1st Qu.: 1.00   N:6271447          1st Qu.:132.0   1st Qu.:113.0   1st Qu.:1.00    1st Qu.:    6.50  
 Median :2.00    2020-01-09 18:13:00:     20   2020-01-03 00:00:00:     59   Median :1.00    Median :     1.60   Median : 1.00   Y:  68120          Median :162.0   Median :162.0   Median :1.00    Median :    9.00  
 Mean   :1.67    2020-01-09 07:51:00:     18   2020-01-02 00:00:00:     58   Mean   :1.52    Mean   :     2.93   Mean   : 1.06                      Mean   :164.7   Mean   :162.7   Mean   :1.27    Mean   :   12.69  
 3rd Qu.:2.00    2020-01-09 08:15:00:     18   2020-01-06 00:00:00:     58   3rd Qu.:2.00    3rd Qu.:     2.93   3rd Qu.: 1.00                      3rd Qu.:234.0   3rd Qu.:234.0   3rd Qu.:2.00    3rd Qu.:   14.00  
 Max.   :2.00    2020-01-10 07:54:00:     18   2020-01-16 00:00:00:     56   Max.   :9.00    Max.   :210240.07   Max.   :99.00                      Max.   :265.0   Max.   :265.0   Max.   :5.00    Max.   : 4265.00  
 NA's   :65441   (Other)            :6404893   (Other)            :6404641   NA's   :65441                       NA's   :65441                                                      NA's   :65441                     
     extra            mta_tax          tip_amount        tolls_amount      improvement_surcharge  total_amount      congestion_surcharge
 Min.   :-27.000   Min.   :-0.5000   Min.   : -91.000   Min.   :-35.7400   Min.   :-0.300        Min.   :-1242.30   Min.   :-2.500      
 1st Qu.:  0.000   1st Qu.: 0.5000   1st Qu.:   0.000   1st Qu.:  0.0000   1st Qu.: 0.300        1st Qu.:   11.16   1st Qu.: 2.500      
 Median :  0.500   Median : 0.5000   Median :   1.950   Median :  0.0000   Median : 0.300        Median :   14.30   Median : 2.500      
 Mean   :  1.115   Mean   : 0.4923   Mean   :   2.189   Mean   :  0.3488   Mean   : 0.298        Mean   :   18.66   Mean   : 2.276      
 3rd Qu.:  2.500   3rd Qu.: 0.5000   3rd Qu.:   2.860   3rd Qu.:  0.0000   3rd Qu.: 0.300        3rd Qu.:   19.80   3rd Qu.: 2.500      
 Max.   :113.010   Max.   :30.8000   Max.   :1100.000   Max.   :910.5000   Max.   : 0.300        Max.   : 4268.30   Max.   : 2.750      
                                                                                                                                        

dfNew <- df %>%
  filter(tip_amount > 0 & tip_amount < 500)